Morphological, physico-biochemical and marker-based diversity of desi cotton (Gossypium herbaceum L.) germplasm

Abstract

Desi cotton (Gossypium herbaceum) is one of the important crops as it is valuable source of lint and spinnable fibre for textile industries. G. herbaceum is resistant to biotic and abiotic stresses and is sturdy crops species. The genetic diversity analysis of this crop is important for further improvement in its productivity to cap the gap between demand and supply of short staple cotton. Therefore, a set of 48 inbred lines of desi cotton were used in this study and their variability was estimated using morphological traits, yield parameters, seed physical properties, fibre quality parameters and seed chemical parameters. Similarly, 13 SSR and five ISSR markers were used for molecular diversity evaluation of germplasm. In this study, ANOVA showed significant differences among all genotypes for all the traits, demonstrating a substantial amount of genetic variability in studied genotypes. Morphological studies showed that genotype Radhanpur had higher seed yield (185.50 g/plant) and lint yield (68.60 g/plant). Physico-biochemical studies suggest that genotypes GVhv-845 had higher fibre strength (26.10 g/tex), Wagad (19.05%) was higher in oil and W8 (51.67%) had maximum seed protein. The Manhattan dissimilarity co-efficient based phenotypic diversity generated six main clusters. The average dissimilarity value among genotypes was 0.30, indicating moderate phenotypic variability. The dendrogram generated from pooled data of SSR and ISSR markers based on Jaccard’s similarity matrix grouped the genotypes into four main clusters. The genetic coefficient of similarity among the genotypes ranged from 0.15 to 0.70 with an average of 0.32. The present study revealed a low correlation between phenotypic and marker-based matrices(r = 0.09). Along with, low correlation, both the matrices placed a limited number of genotypes (nine) in the same clusters in their respective dendrograms. The low correlation indicated that the two methods were different and highly variable as molecular markers are neutral in behave than quantitative traits.